Tag Archives: game history

My Favorite Game Postmortems


This game was confusing as hell to me as a child, but I loved its atmosphere.


I think the main takeaways for Myst are:

  • Rand and Robin Miller made a lot of simpler point-and-click games for children before tackling Myst.
  • Dungeons and Dragons was a large inspiration because of the emergent storytelling aspects. Rand actually designed a dungeon which was later used in creating one of Myst’s worlds.
  • The game took its mythic structure from Star Wars, its world-traveling from C.S. Lewis’s The Chronicles of Narnia, and the island setting was inspired by novels by Jules Verne.
  • Myst was created for non-gamers, so they tried to avoid arbitrary puzzles that most adventure games suffered from in the day. They mostly succeeded, though Robin admits there are still a few bad puzzles in the game.
  • The two brothers approached Myst as a passion project strictly for themselves.
  • How long it took to render each image back in the day. Shots took 2-14 hours to render using a StrataVision 3D.
  • They subverted the one-way communication of the story with one-way devices like TVs and books rather than personal encounters.
  • They initially used diegetic music in their games (music that only occurs naturally in the world). When they finally added traditional soundtrack music to Myst, they were surprised how well it fit.
  • Extensive beta testing helped make the game playable to non-gamers.
  • Myst and their previous games were all made on HyperCard.
  • Myst took two years to complete.


Also, Phil Fish makes a surprising appearance 51 minutes in. Apparently he started out on HyperCard as well.



The original ARPG, and one of my favorite games of all time.


I think the main takeaways for Diablo are:

  • The interesting meeting between Condor Games (Diablo’s developers) and Silicon & Synapse (Warcraft’s developers). The latter became Blizzard Entertainment. The former became Blizzard North.
  • The tiles in Diablo were borrowed directly from X-COM.
  • How they handled drawing the game with only 256 colors.
  • David Brevik had never coded in C before Diablo.
  • All the interesting bug fixes and work-arounds they had to implement because of the technology at the time.
  • There was a small bidding war between 3DO and Blizzard for Condor Games. Condor went with Blizzard despite 3DO offering twice as much money.
  • Most of the HUD (graphical interface) work was done in the last 3 months, with the help of input from beta testers.
  • They crunched really hard on this game for the better half of a year. Brevik says crunch deserves its bad reputation, but is also a “necessary evil” and not entirely a bad thing. From my own personal experience, I can definitely relate.
  • There’s a spell called “Blood Exchange” they removed from the game. It let you swap your HP with a monster’s HP, making the game really easy to beat.
  • Rogue, NetHack, Moria, Angband, Ultima, Wizardry, Doom, Dark Forces, and X-COM were just some of the influences for Diablo.
  • battle.net ran on one computer! It handled the entire multiplayer side of the game.


I also love the Ms. Pac Man and Marble Madness postmortems, but they’re literally so old (arcade era) that the advice isn’t really applicable to modern game design. They’re more like historic documents now. Which may be the biggest takeaway of all…


Emergent Gameplay #2, No Man’s Sky and Holy Grails

No Man’s Alibi

As far as I can tell, No Man’s Sky does not have emergent gameplay. Nothing exciting is going to happen as a result of combining these procedural generation engines. You won’t even be able to have emergent multiplayer experiences, because Sean Murray and Hello Games have clearly lied about multiplayer being an available feature.

Not only that, but they avoided denying it by saying the odds of meeting someone in the game were virtually impossible due to its sheer size. Two lets players managed to meet up on PS4 the first day the game was released. So that was another lie.

You hear animal noises on planets with no animal life. You cannot terraform the planets in any meaningful capacity. The game frequently crashes. No unique ships. No factions. No intelligent aliens outside of buildings, ever. Reaching the center of the universe, the only real “quest”, is hugely disappointing. 18 quintillion planets of emptiness and sameness. The list of removed features, broken promises and issues have been cataloged on reddit, and it is overwhelming.

But what really kills me is they haven’t learned anything from the current situation. Sean Murray is still actively promising base building and “giant space freighters” in a future update – once they’re done handling customer support and fixing crashes on PS4. They cannot stop getting ahead of themselves. They are still trying to have their cake and eat it too.

Worst of all, they’re still using trailers unrepresentative of the game on Steam and their website in order to sell their product. Steam, for their part, are accepting refunds from users with more than two hours of gameplay. The game’s sales have already dropped by 81% in its second week. Right now, No Man’s Sky looks more like an alpha demo than a finished product.

I definitely think Sean Murray is making similar mistakes to Peter Molyneux, as far as dreaming too much while marketing his game.


Emergent Gameplay and Glitches

For now, the Holy Grail of emergent game design is still just a selling point for companies like Hello Games. The promise of truly emergent game design still has not been achieved. Maybe it won’t be for a long time.

However, what we can do right now is look at No Man’s Sky and where it failed, and think about how game developers can do better in the future.

When people go on about the importance of randomness and emergence in video games, they are usually referring to the factor of unpredictability it brings. If your game is highly unpredictable, then you have implemented procedural generation well. However if your game is too predictable then you could argue that the procedural generation could have been implemented better. Sometimes while trying to smooth bugs out of the game, the latter scenario can happen.

A potential paradox here is that glitches are a good source for emergent gameplay. If you completely iron out the glitches in your procedural generation, you could be losing potentially intriguing aspects of the game. The Far Lands in Minecraft is a good example. Glitches occur when you travel to extremely far distances, but you also find unique structures that wouldn’t occur anywhere else. These glitches can add an extra level of replayability.

Of course a distinction should be made between glitches that are difficult to find and exploit, vs bugs that occur on a regular basis and have to be removed for a polished game. Sometimes, especially from a developer’s standpoint, the distinction between the two cannot be all that clear.

The question is does a AAA game have room for that kind of experimentation? In the case of No Man’s Sky, clearly not.

Hopefully in the future, games media will be more pressing about the specifics of game development. I think the media’s “we’ll take your word for it” attitude is partially to blame for the outrage surrounding the game. They report Sean’s non-explanations as explanations and continue defending the game, basically to avoid the wrath of the hostile NMS fanbase.


Holy Grails

I first wrote about No Man’s Sky back in January 2015 in an article about emergent gameplay. Replacing pre-written dialogue with chatbot AI, dynamic voice synthesis to create unique voices for NPCs, expressing complex emotions without dialogue, were a few of the ideas I proposed for the potential future of open world games.

When people talk about the Holy Grail of game design, they can be referring to several different ideas. What are the Holy Grails of AI and game design? For a long time it was chess AI, then more complex games like Go, which was only recently conquered by AlphaGo. Infinite, randomly procedurally generated worlds are the current trend in the gaming world. Unlimited replayability is one of the ultimate goals here.

For an idea to qualify as a “Holy Grail”, it has to be an idea easily recognizable to most people but very difficult to attain in practice.

Complete immersion within a video game is considered a Holy Grail to some – a game that never breaks your suspension of disbelief. Highly emergent gameplay with unpredictable outcomes is also regarded as a Holy Grail.

Other realms include the Holy Grail of algorithmic music composition, which has been led by David Cope since 1981. Artificially intelligent replication of famous classical composers has slowly gained popularity in more recent years. On the other side you have programs which compose entirely original music.

Then you have story generator algorithms such as Sheldon Klein’s Novel Writer system from 1973, which generated 2100 word murder stories. Newer algorithms such as MEXICA, BRUTUS, and FABULIST have pioneered increasingly complex methods to generate stories. Algorithms for poetry are still uncommon, but I suspect that will change in the future.

For many, chatbot AI that can pass the Turing test is the Holy Grail.

To many others, the Holy Grail is simply generalized AI. Or more specifically the “singularity” event, where generalized AI begins exponentially acquiring more intelligence.


The Outsider

So which Holy Grail am I referring to? I’m referring to David Braben (creator of Elite) and his use of the term in 2005, when talking about a new golden age of gaming. He spoke of storytelling in video games breaking free of their “non-interactive roots”.

In games on the fifth generation, glorious imagery is not a problem. Details like beads of sweat are now expected.

Natural-looking fur can make creatures and people look very real, though non crew-cut hair brings many additional problems if it is not set in place with military strength hair gel, but that is not the point.

This is merely a slight increment over what fourth generation games are doing on the PlayStation 2, Xbox and GameCube and is behind the complaints being voiced – the appetite is there for more and we are tired of the car-on-a-train-track spectacle.

Story-telling in games in most cases is little different to the stories of those Harold Lloyd films of the 1920s.

The player is stuck on pre-defined railway lines, forced to follow their character’s pre-determined adventures, much as in a book or a film.

In story-telling terms at least, games have not yet broken free of their non-interactive roots.

The Holy Grail we are looking for in fifth generation gaming is the ability to have freedom, and to have truly open ended stories.

Games that have even hinted at that freedom in the past like Elite and Grand Theft Auto have been hugely successful. This Holy Grail is what will herald the new era for gaming.”

Reading this again, it sounds a lot like Sean Murray gushing over the (unrealized) possibilities of No Man’s Sky, procedurally generating every fiber and atom in the game and so forth. Of course, you have to shoot for the moon if you want to hit the stars.

During that 2005 interview and for a few years afterwards, David kept hyping up Frontier’s game The Outsider as the first true fifth generation game. Wonder why you’ve never heard of it? It was cancelled around January 2011. It seems they had trouble finding a publisher, and EA wanted to turn it into a game for the Bourne franchise.

David re-confirmed the cancellation of the game in 2014. Yet another example of a potentially great game being shelved due to executive decisions, a shifting market, and an unwillingness to take risks. How many executive decisions affected No Man’s Sky development, I wonder? Here is another enlightening interview, discussing innovation and interactive narratives in video games.

So why do you think story hasn’t moved forward as much? Do you think you have the magic bullet that no one else has discovered?

DB: No one has a magic bullet. We’ve been working on it a long time. It’s a risk as well. The reason I mention The Darkness and BioShock is, story’s not a priority in reviewers’ eyes, and I think that’s a sad thing. […] We really need to move forward on story — as one of the fronts. That’s not the only front left. Look at Katamari. A while before that, people were saying there are only so many genres, and then someone comes out with a lovely game. And tomorrow it will be something similar.


I had an argument with somebody that there were only four types of gameplay, and then out comes Populous. Okay, there are five, then. And usually, it’s an excuse to plagiarize. We all take inspiration from other games, and that’s fine. It’s when we take inspiration and don’t do any more. That’s the sad thing. When you don’t move it forward. And there’s a danger. […] Especially these days, where we’re making fewer games than we used to. We’re essentially being trusted to use the opportunity to do something fantastic, and if we don’t we should get slapped around — which I’m sure we will do. [Laughs.] – Citation

So what point am I trying to make? The point is most game developers are not in an environment that allows for the pursuit of experimental ideas, lofty thoughts, or “Holy Grails”. Trying to break new ground is risky, and scares investors. So the big publishers play it safe and settle for mediocrity. That’s why the overly ambitious No Man’s Sky came from a team of 15 people, and why Minecraft was started by a single person.

For now, it seems that true innovation still has to come from outside the AAA video game industry.

The “Story-Driven Game”

Mojang pocketed 2.5 billion dollars with Minecraft. In a sense, this is a huge victory for procedural generation, which has taken a backseat to RPGs and story-driven games over the recent years. And so I am sorting through my ideas, figuring out what I want to keep and what I want to abandon.

There was a time when making a successful indie game with a story seemed feasible. Cave Story was released in 2004 and became very popular, around the time I discovered Game Maker. Were Cave Story released now, I doubt many people would even see it, much less play it. And the games I considered popular among the GM scene – Jumper 2, Seiklus, The Legend of Shadow – would be entirely drowned out on Steam or Google Play.

I was (re)inspired to make one of these games when I played Kingdom Hearts for the first time in 2011. But when I began trying to write stories for my games, I learned how difficult they fit together. Story has to bend to fit the gameplay, and vice versa. Games bring elements that aren’t experienced in other media, like real-time interaction. Simply firing the trigger in a game lowers activity in the amygdala, which processes how we manage emotions. This clashes when you have a heart-wrenching story you’re trying to tell.

Tetsuya Nomura’s work in Final Fantasy and Kingdom Hearts is a good example of where this gets confusing. He is a game designer, but he is also a “director”, which is an unusual term in gaming.

Final Fantasy XV, which he’s leaving for other projects, is no doubt beautiful and has great-looking gameplay, but in the trailer that takes a backseat (no pun intended) to the cutscenes. The Final Fantasy series showcases a lot of things that don’t naturally occur in games. Dialogue, character relationships, serious emotional situations, plot twists, the story’s effect on the “character” rather than the “player”, and so on.

Here is a quick history lesson. In the 80s and 90s, graphic adventures such as point-and-clicks were very popular on PC thanks to LucasArts and Sierra Entertainment. Cyan World’s Myst became the best selling PC game in 1994 (until The Sims in 2002). These games had rich graphics and engrossing storylines, but difficult interfaces and puzzles.

Then something happened on June 23, 1996. It was called the Nintendo 64. More specifically Super Mario 64, which gave us full control over a character in a 3D world. Playing in 3D games had an intoxicating novelty back then. By the time Myst’s sequel Riven came out in 1997 it was ignored, and point-and-clicks felt dated.

Graphic adventures were more interesting in a literary sense, but weren’t as “fun” as interacting in 3D. And so LucasArts and Sierra shut down at the turn of the century, while most console games developed absent of depth beyond “save the princess, kingdom, whatever”.

I get frustrated with the “cutscene, gameplay, cutscene” formula games still follow, as it makes gameplay and story feel detached. But the biggest problem is – linear stories add very little replayability. Unless you truly enjoy the game engine or story, you’re just re-experiencing what you went through before. Linear game-stories have lost their novelty just like graphic adventures before them.

I started designing games at a strange time, looking back. Indie games could be counted in the thousands then – not millions. There were no indie markets, so many of us gave our games away. I can’t recall a single indie developer selling their games ten years ago. Even Cave Story was free. We all survived on word of mouth, not paid advertising.

Another history lesson. I’ll use poetry as an example. This is the sixth entry from Sylvia Plath’s diaries: “Today is the first of August. It is hot, steamy and wet. It is raining. I am tempted to write a poem. But I remember what it said on one rejection slip: ‘After a heavy rainfall, poems titled ‘Rain’ pour in from across the nation.'”

This is how I feel indie games have become, in the extremely short span of ten years. Most of us are just writing rain poems now, to be lost in an ocean of other indie games.

Ludum Dare has become very popular. Game jams are everywhere now – even the White House ran one. To me, this is further indication that game design itself has become something casual, like writing poems or music.

The problem is that “game design” wasn’t originally an art form where one person has 100% creative control. In this sense, games do bear more similarity to movies than visual art, poetry, or music. And there aren’t many movies made by one person. While game development is becoming more of a hobby, games are still multimedia, and flourish from the synergy of people working together, sharing their ideas.